1. Field of the Invention
The present invention relates to a semiconductor integrated circuit device intended to reduce its power consumption.
2. Description of the Prior Art
Recent developments in portable or mobile equipment have caused requirements of low power semiconductor integrated circuit devices, which can be built in mobile equipment, for long battery life. Reduction in the operating voltage of a semiconductor integrated circuit device is one of effective methods of reducing the power consumption. Since the power consumption is the product of the current and the voltage, both the voltage and the current can be reduced along with the reduction in the operation voltage. It is therefore said that in general, reduction in the operating voltage of a semiconductor integrated circuit device can produce a squared effect with respect to the reduction in the power consumption.
On the other hand, a MOSFET that composes a part of a semiconductor integrated circuit device has the property that, as the power voltage applied thereto is reduced, its performance is reduced and hence its response speed is reduced. To avoid the reduction in the performance of a MOSFET, it is necessary to reduce the threshold voltage according to the reduction in the power voltage. However, the threshold voltage cannot be reduced in proportion to the reduction in the power voltage, in practice. This is because the reduction in the threshold voltage causes an increase in the amount of a leak current flowing in a MOSFET when the MOSFET is held in its off state and hence an increase in the power consumption.
To overcome the above problem, a prior art method as disclosed in Japanese Patent Application Laying Open (KOKAI) No. 7-212218 has been used. Referring next to FIG. 17, there is illustrated a schematic circuit diagram of a prior art semiconductor integrated circuit device which can run from low voltage supply. The device is constructed from a multi-threshold CMOS or MT-CMOS. In the figure, reference numeral 1 denotes a two-input NAND gate, 2 and 3 denote p-channel MOSFETs, 4 and 5 denote n-channel MOSFETS, 6 denotes a p-channel MOSFET which can be turned on in response to a control signal 9 asserted LOW so as to connect a power supply 7 with a power supply line 8 (VA1), and 10 denotes an n-channel MOSFET which can be turned on in response to a control signal 13 asserted HIGH so as to connect a ground potential node or GND 11 with a power supply line 12 (VB1). The absolute values of the threshold voltages of the p-channel MOSFET 6 and the n-channel MOSFET 10 are set to be greater than those of the threshold voltages of the p-channel MOSFETs 2 and 3 and the n-channel MOSFETs 4 and 5 which construct the two-input NAND gate 1.
Each of the p-channel MOSFETs 2 and 3 has a substrate terminal connected to the VA18, and each of the n-channel MOSFETs 4 and 5 has a substrate terminal connected to the VB112. The p-channel MOSFET 6 has a substrate terminal connected to the power supply 7, and the n-channel MOSFET 10 has a substrate terminal connected to the GND 11.
When activating the two-input NAND gate 1 as shown in FIG. 17, the control signal 9 is caused to make a HIGH to LOW transition and the control signal 13 which is the inversion of the control signal 9 is therefore caused to make a LOW to HIGH transition. Both the p-channel MOSFET 6 and the n-channel MOSFET 10 are then turned on, and therefore the potential of the VA18 rises to the potential of the power supply 7 and the potential of the VB112 descends to the level of the GND 11. Consequently, the two-input NAND gate 1 can operate as a normal NAND circuit. In this case, since the absolute values of the threshold voltages of the MOSFETs 2 through 5 are set to be relatively small, the NAND gate 1 runs at a high speed even though the voltage from the power supply 7 is low.
When deactivating the two-input NAND gate 1, the control signal 9 is caused to make a LOW to HIGH transition and the control signal 13 which is the inversion of the control signal 9 is therefore caused to make a HIGH to LOW transition. Both the p-channel MOSFET 6 and the n-channel MOSFET 10 are then turned off and therefore the VA18 and the VB112 are disconnected from the power supply 7 and the GND 11, respectively. Since the absolute values of the threshold voltages of the p-channel MOSFET 6 and the n-channel MOSFET 10 are set to be greater than those of the threshold voltages of the p-channel MOSFETs 2 and 3 and the n-channel MOSFETs 4 and 5, the amounts of leak currents flowing through the MOSFETs in the two-input NAND gate 1 can be controlled to small ones.
In general, when the voltage between the gate and the source of a MOSFET is less than or equal to its threshold voltage, the leak current that flows between the source and the gate of the MOSFET increases exponentially with an increase in the gate voltage. Therefore, making a difference between the threshold voltages of the MOSFETs 2 through 5 and those of the MOSFETs 6 and 10 can reduce the leak currents flowing through the two-input NAND gate 1 not in active use by a large amount. A similar explanation can be made for any other type of internal circuitry having any size such as a logic circuit or a storage circuit, included in a semiconductor integrated circuit device, other than the two-input NAND gate 1 shown as an example of the internal circuitry.
A problem with such a prior art semiconductor integrated circuit device which can run from low voltage supply is that when the internal circuitry included in the semiconductor integrated circuit device is a combinational circuit in which its output is determined by a combination of inputs, such as a two-input NAND gate as mentioned above, the internal circuitry works properly, but, when the internal circuitry is a sequential circuit such as a latch circuit in which its output depends on previous input conditions, there is a possibility that the internal circuitry fails to work properly, as will be explained below.
Referring next to FIG. 18, there is illustrated a schematic circuit diagram showing another prior art semiconductor integrated circuit device including internal circuitry which serves as a sequential circuit. The semiconductor integrated circuit device is so-called a latch circuit in which the inputs and outputs of two inverters are connected to each other while the connections cross each other. In the figure, reference numeral 20 denotes a latch circuit, 21 and 22 denote p-channel MOSFETs, and 23 and 24 denote n-channel MOSFETs. The other structure of the semiconductor integrated circuit device is the same as that of the semiconductor integrated circuit device shown in FIG. 17. The absolute values of the threshold voltages of the p-channel MOSFET 6 and the n-channel MOSFET 10 are set to be greater than those of the p-channel MOSFETs 21 and 22 and the n-channel MOSFETs 23 and 24 which construct the latch circuit 20.
As shown in FIG. 18, the latch circuit 20 is constructed of the p-channel MOSFETs 21 and 22 and the n-channel MOSFETs 23 and 24 having their threshold voltages which are relatively small. A node 25 and a node 26 pair up with each other. When one of the pair of nodes 25 and 26 is held in a HIGH logic level, the other one of the pair is held in a LOW logic level.
When a control signal 9 asserted LOW and a control signal asserted HIGH are applied to the p-channel MOSFET 6 and the n-channel MOSFET 10, respectively, the latch circuit 20 is activated. Then the nodes 25 and 26 can hold values normally. Data can be written into or read out of the nodes 25 and 26 at a high speed because the absolute values of the threshold voltages of the p-channel MOSFETs 21 and 22 and the n-channel MOSFETs 23 and 24 are relatively small. However, if an attempt is made to reduce leakage currents flowing through the MOSFETs when the latch circuit 20 is deactivated in response to the control signal 9 at a HIGH logic level and the control signal 13 at a LOW logic level, data cannot be held at the nodes 25 and 26 because leakage currents flowing through the p-channel MOSFETs 21 and 22 and the n-channel MOSFETs 23 and 24 when the latch circuit 20 is deactivated become greater than those flowing through the p-channel MOSFET 6 and the n-channel MOSFET 10 when the latch circuit 20 is deactivated. For example, in a case where the node 25 has a HIGH logic level and the node 26 has a LOW logic level, the p-channel MOSFET 22 and the n-channel MOSFET 23 are turned off and the p-channel MOSFET 21 and the n-channel MOSFET 24 are turned on, then the potential of the node 25 at a HIGH logic level descends and the potential of the node 26 at a LOW logic level rises because of leakage currents flowing through the p-channel MOSFET 22 and the n-channel MOSFET 23 until they are equal to each other, and finally the data held at the nodes 25 and 26 are erased.
As previously mentioned, a problem with such a prior art semiconductor integrated circuit device is that when a latch circuit is disposed as an internal circuit of the semiconductor integrated circuit device, data held at nodes within the latch circuit are erased because of leakage currents flowing through MOSFETs which construct the latch circuit.